JPH03179838A - Signal detection system - Google Patents

Abstract

PURPOSE:To prevent the deterioration of a detector characteristics for mis- detection and to prevent the deterioration in the transmission line efficiency without changing the length of a synchronizing signal by using power between NT of a soft discrimination data string so as to correct an identification value. CONSTITUTION:The system is provided with power averaging circuits 1, 2 receiving a synchronizing demodulation soft discrimination string of a modulation signal with synchronizing words of predetermined number and length inserted periodically therein and obtaining average power per bit number of the predetermined number and a correlation circuit 4 receiving a synchronizing demodulation soft discrimination string and applying the correlation calculation with the pattern of the synchronizing word. Moreover, an identification value setting circuit 3 multiplying a predetermined coefficient with an output signal of the power average circuits 1, 2 to set the identification value and a comparator circuit 5 comparing the output signal with an output signal from the correlation circuit 4 to detect the synchronizing word are provided. Thus, the deterioration in the undetection probability is prevented without varying the length of the synchronizing signal and the deterioration in the transmission line efficiency is prevented.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a signal detection method, and is particularly applicable to a system such as a mobile satellite communication system where signal interference occurs frequently due to rapid fluctuations in the transmission path. The present invention relates to a signal detection method.

[Conventional technology]

Conventionally, this type of signal detection method has determined the presence or absence of a signal by performing correlation detection on a synchronization signal included in the signal.

[Problem to be solved by the invention]

The conventional signal detection method described above detects the correlation of the synchronization signal in the binary decoded signal sequence after hard decision, and if the number of errors is within a certain value, it is determined that the signal is present, and if the number of errors exceeds the certain value, it is determined that the signal is strong. I was judging. The length of the synchronization signal is N, the transmission path error rate is p,
If the above constant number of errors is ε, the non-detection probability Pm1s
s is denoted by. By the way, in a system such as a mobile satellite communication system where the transmission path conditions change greatly, sometimes it becomes impossible to receive a signal at all, and only noise is received. At this time, the error rate of the demodulated sequence is considered to be 1/2 of that of the synchronization word, so the probability Pt of erroneously determining that there is a signal is as follows. As is clear from equation (2), since P decreases in intensity due to -, it is not possible to keep P low unless N is made sufficiently long. Therefore, there has been a drawback that the efficiency of the transmission path is lowered.

[Means to solve the problem]

The signal detection method of the present invention receives as input a synchronous demodulation soft decision sequence of a modulated signal in which synchronization words of a predetermined number of lengths are periodically inserted into the signal, and the predetermined a power averaging circuit that calculates the average power per bit number of the number of bits; a range value setting circuit that takes the synchronous demodulation soft decision sequence as input and sets a threshold value by multiplying the synchronous word by the number of bits; and this range value setting circuit. and a comparison circuit for detecting a synchronization word by comparing the output signal of the correlation circuit with the output signal of the correlation circuit.

In the signal detection method of the present invention, a squaring circuit that squares the synchronous demodulation soft decision sequence and an averaging circuit that averages the output signal of the squaring circuit in an interval of the predetermined number of bits A power averaging circuit may also be configured.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention.

An input data sequence 101, which is a synchronous demodulation soft decision sequence, is input to a squaring circuit l, and γ2(t) is output. Here γ(1
) is the input data series 101.

The squared sequence is processed by the following averaging circuit 2 to obtain the synchronization word length N
The bits are averaged over the interval. From this, the average power of the input data series 101 at each time is determined.

The output signal of the averaging circuit 2 is applied to one input of the comparison circuit 5 via the occupational value setting circuit 3. On the other hand, the input data series 101 is also input to the cross-correlation circuit 4.

The cross-correlation circuit 4 performs the following calculations. However, y(t) is the synchronization word itself and is zero outside the interval (-NT/2.NT/2),
* indicates complex conjugation. If a synchronizing signal y(t) is periodically inserted into γ(1), its output will be combed at a certain point in time. γ(1) can be written as γ(t)=y(t)+n(t)(5) at that point. Here, n(ri is a noise signal. Substituting equation (5) into equation (4), we get n (t) to 6,000 impulse responses y“
(This is the output signal when passed through the −〇 filter.

Therefore, it can be seen that the cross-correlation circuit 4 improves the signal-to-noise power ratio by N times. Therefore, if N is large enough, the second
Most of the items can be ignored. For example, if N is 10, S/N of input data series 101 = y 2/
In the output of this cross-correlation circuit 4, n' is S/N==
10 y '/n 2, and the noise power is l/10
It becomes equivalent to .

Since the output signal Z(+) of the averaging circuit 2 is, substituting equation (5) gives the following equation. The third term in equation (8) is y (t) and n (t)
are strongly correlated with each other, and the autocorrelation function of n(+) is N
If it converges quickly enough for T, it follows from the ergodic theorem. And if the average value of n (t) and y (t) is zero, the third term is F! It can be seen as one town. Furthermore, the second
The term is usually equal to or smaller than the first term (S/N≧
00 conditions).

Now, the knowledge range value ε obtained from Z (t) is multiplied by the number of threads α by the occupational range value setting circuit 3, and is calculated as ε(t)=α・Z(
t) ...00). This coefficient α is set to a value that reliably detects ψ(0) in equation (6) and prevents detection of ψ′(0), which would result in a pattern other than a synchronized word being U(r), as much as possible. be done. However, in a system like the one discussed here, there is usually a subsequent frame synchronization circuit, so there is no need to consider the case of u (t). A problem arises when y(t) (y(t) in equation (5)) in the input data series 101 becomes almost zero due to attenuation of the transmission path.

If the gain of the demodulator is always constant, there is no need to use an adaptive threshold value, but receiving equipment usually uses an AGC circuit, so γ2(t) = constant...Q
It will be 21. Look at ψ(0) and ε(1) under this condition.

AG is set so that the power when y (t) exists in the input data series 101 is equal to the power when y (t) becomes intense.
Since C operates, for the sake of distinction, let n+(t) be the noise signal when there is no y(t). This formula is further expressed as y (t)'+ n (t)2= n 1(t)2
...04). Here n(
t) and y(t) are assumed to be independent and have an average of 0.

Furthermore, n+(t)2=n(t)2(1+S/N). From this, when coefficient 0 is substituted into equation (6), it can be seen that the noise term has become larger due to the disappearance of y (t) from the input data series 101.

On the other hand, from equation (8), Z (t) for determining the threshold value is obtained by considering the third term as a fish, and it can be seen that a correction is made for the increase in y (t). .

The comparison circuit 5 compares the magnitude of ψ(o), which is the output of the cross-correlation circuit 4, and ε(1), which is the output of the occupational value setting circuit 3, and determines the presence or absence of a synchronization word.

[Effect of selling out]

As explained above, in the present invention, signal detection is performed by cross-correlating demodulated soft-decision data strings. This has the effect of preventing deterioration of the false detection/non-detection characteristic without changing the length of the signal N, and preventing a decrease in transmission path efficiency.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention. l... Square circuit, 2... Average circuit, 3
. . . Criterion value setting circuit, 4 . . . Cross-correlation circuit, 5 . . . Comparison circuit.

Claims (1)

[Claims] 1. A synchronization demodulation soft decision sequence of a modulated signal in which synchronization words of a predetermined length are periodically inserted into the signal is input, and the predetermined method of this synchronization demodulation soft decision sequence is performed. a power averaging circuit that calculates the average power per bit of the number of bits; a cross-correlation circuit that receives the synchronous demodulation soft decision sequence and performs a cross-correlation operation with the pattern of the synchronous word; and an output signal of the power averaging circuit. A threshold value setting circuit that sets a threshold value by multiplying it by a predetermined coefficient, and a comparison circuit that compares the output signal of the threshold value setting circuit with the output signal of the correlation circuit to detect a synchronization word. This signal detection method is characterized by: 2. The power averaging circuit is configured by a squaring circuit that squares the synchronous demodulation soft decision sequence, and an averaging circuit that averages the output signal of the squaring circuit over an interval of the predetermined number of bits. The signal detection method according to claim 1, characterized in that: